Maria Angelena

AI Artist & Digital Creator
Creative Technologist. Sharing ideas, prompts, workflows and things I craft.

[𝗦𝗮𝗻𝗶𝗮 @saniaspeaks_]

What if an AI girl traveled to China to learn Kung Fu?

Seedance 2.5 via higgsfield makes it feel real.

Prompt:
Smartphone sports vlog featuring one Korean woman in her early 20s learning and practicing traditional Chinese martial arts (kung fu) at an authentic outdoor training courtyard in China. The entire video must look like genuine casual smartphone footage, with imperfect handheld framing, subtle camera shake, realistic autofocus adjustments, automatic exposure changes, slight motion blur, natural walking movement, realistic skin texture, authentic ambient sounds and real-world physics. No cinematic shots, no gimbal, no drone, no beauty filters, no commercial look.

Character: The same Korean woman throughout the entire video, petite oval face, brown eyes, natural coral lips, realistic fair skin, long black hair. Maintain perfect character consistency with no face morphing, identity drift or duplicate people.

0:00–0:04 — Training Arrival
She stands outside a traditional Chinese martial arts training courtyard in selfie mode, looking excited. She says in Korean: “오늘 중국 무술을 배워볼 거예요!” She turns the phone to show the training area and students practicing in the background.

0:04–0:08 — Getting Ready
She changes into a simple traditional-style kung fu training uniform. She stands beside an instructor and carefully copies basic martial arts movements, looking focused but slightly nervous.

0:08–0:12 — First Practice
She practices a simple sequence of controlled punches and defensive movements. Her movements are believable for a beginner, with small mistakes and natural corrections rather than perfect professional technique.

0:12–0:15 — Kick Practice
She tries a controlled front kick toward a training pad held by the instructor. She slightly loses balance afterward, laughs naturally and looks embarrassed.

0:15–0:18 — Hard Cut / Improved Practice
Hard cut exactly at 0:15. Same woman, same identity and training outfit. She now performs the same basic movements with noticeably better confidence and coordination.

0:18–0:22 — Martial Arts Sequence
She performs a short, simple kung fu combination with controlled punches, blocks and footwork. Keep all movements physically realistic, safe and believable.

0:22–0:26 — Training With Others
She joins several students practicing synchronized basic movements in the courtyard. Natural sounds of footsteps, clothing movement and distant conversations are audible.

0:26–0:30 — Happy Ending
Breathing slightly heavily after training, she returns to selfie mode, smiles proudly and says: “생각보다 어렵지만 정말 재미있었어요!” She gives a small wave and naturally lowers the phone to end the vlog.

Throughout the entire video, prioritize authentic smartphone realism: imperfect framing, subtle focus changes, realistic breathing, natural blinking, believable beginner movements, physically accurate body motion, authentic courtyard ambience and spontaneous reactions. The final result should feel like a real personal vlog filmed casually on a smartphone, not an AI-generated action movie or commercial.

[Lyra AI @AI_with_Lyra]

JOVAN White Musk luxury in every drop

Prompt:

Create a premium ultra-photorealistic live-action luxury fragrance commercial, approximately 10 seconds, vertical 9:16 aspect ratio, designed with high-end cinematic product advertising aesthetics.

SCENE & PRODUCT: A JOVAN WHITE MUSK fragrance bottle stands upright on a glossy black marble surface, with its matching retail box positioned behind it. The transparent bottle contains amber-colored perfume and features a silver metallic ribbed cap with a clean white label. The composition feels elegant, luxurious, sophisticated, and premium.

PRODUCT LOCK — CRITICAL: Preserve the exact original JOVAN WHITE MUSK bottle and packaging design. Keep the logo, WHITE MUSK text, typography, label layout, bottle proportions, amber liquid, silver ribbed cap, colors, packaging, and branding exactly as shown. No redesign, distortion, missing text, or altered logos.

OPENING SHOT: Begin with a cinematic close-to-medium hero shot of the bottle and box on reflective black marble. The background is dark and softly blurred, featuring subtle botanical elements, warm golden bokeh, atmospheric particles, and gentle light rays.

CAMERA MOVEMENT: Use a slow cinematic push-in with shallow depth of field, keeping the product and label sharply focused. Transition smoothly into an extreme macro shot of the atomizer.

ACTION: The atomizer is pressed and releases a fine perfume mist in slow motion. Capture the tiny droplets with warm backlighting and realistic detail.

TRANSITION: As the mist spreads, smoothly transition back to the full product shot. Glowing golden perfume bubbles begin floating naturally around the bottle with realistic reflections and refraction.

LIGHTING: Use warm golden rim lighting, soft studio illumination, realistic highlights on the glass and metallic cap, deep cinematic shadows, subtle volumetric rays, and elegant reflections on the marble.

CAMERA & OPTICS: High-end cinema camera, macro product photography, 85mm/100mm macro lens look, shallow depth of field, smooth focus pulls, natural bokeh, realistic motion blur, HDR, and detailed glass and liquid textures.

FINAL SHOT: End with a centered hero shot of the JOVAN WHITE MUSK bottle surrounded by golden bubbles and subtle atmospheric particles. Hold on the product with a luxurious cinematic finish.

VISUAL QUALITY & MOOD: Ultra-photorealistic luxury advertising, physically accurate glass and liquid, realistic perfume mist, premium cinematic color grading, smooth motion, crisp details, elegant, sensual, mysterious, sophisticated, and high-end.

[AIwithMinal @AIwithMinal]

Strength, courage, and determination collide as the battle intensifies, proving that true warriors never back down when the challenge gets tough.

Created with seedance 2.0

Prompt:

Ultra-realistic cinematic martial arts scene inside a sunlit school gym, young fighters in dark athletic outfits engaged in an intense choreographed battle, dynamic dodging and striking poses, wooden basketball court, warm golden sunlight streaming through large windows, dramatic dust particles, realistic motion blur, intense expressions, cinematic handheld camera movement, shallow depth of field, natural shadows, high-detail environment, action-movie atmosphere, photorealistic 4K, dramatic cinematic color grading, 16:9.

[Avelyrah @AvelyrahnAI]

Drop the maid outfit, unleash the action. ⚡💼

Created with Seedance 2.5

Prompt:

A fierce female secret agent with white hair styled in a bun, wearing a grey maid uniform and a white face mask, runs at full speed down a luxurious dark marble corridor with warm vertical LED strip lighting and floor-to-ceiling glass windows overlooking a moonlit ocean at night, tightly clutching an antique wooden box. Two burly bodyguards in black suits chase her. She leaps into the air and delivers a powerful flying side-kick into the chest of the lead guard, sending him crashing backward. She lands gracefully, turns sharply, kneels to place the wooden box on the floor, and stands up. In a dramatic motion, she pulls off her white face mask and flings it away, tearing off her maid uniform jacket to reveal a sleek white athletic crop top and shorts. She engages in high-speed martial arts combat, leaping into mid-air splits, dodging sword strikes, and executing dynamic counter-attacks against multiple armed opponents in a flurry of motion.

[Sharon Riley @Just_sharon7]

AI finally gets outfit transformations right.

Made by using GPT Image 2 + Seedance 2.0 on @FishCreativeHQ

prompt

Image 1 SHARON'S REAL FACE — sole source of her features, 100% match, zero drift: soft rounded oval face, warm light-beige skin with tiny freckles and beauty marks, GREEN eyes, dense straight feathered brows, small neat nose, full pink lips, small silver drop earring.

Use the uploaded REFERENCE VIDEO Video 1 as the absolute master for timing, camera movement, body motion, pose, facial-expression timing, interaction, transition rhythm, pixel effects, outfit-change effects, framing, and overall choreography.

Replace the original woman with Image 2 as the ONLY replacement human identity.

From this point onward, refer to her only as THE FEMALE CHARACTER Image 1.

CRITICAL REFERENCE PRIORITY:

REFERENCE VIDEO controls:

exact motion

exact timing

exact camera movement

body choreography

pose changes

head movement

hand movement

facial-expression timing

interaction with the other man

transition timing

outfit-change timing

pixel effects

digital clothing-change effects

framing

speed ramps

movement through the pixels

overall visual rhythm

THE FEMALE CHARACTER controls:

identity @Image 2

face

hairstyle

head shape

skin tone

body proportions

gender presentation

The original woman is NOT an identity reference.

She controls motion only.

The generated video must follow the reference video as closely as possible, but the woman must be completely replaced by the same female character @Image 1 throughout the entire sequence.

==================================================

ABSOLUTE IDENTITY LOCK

There is only ONE replacement identity throughout the video:

THE SAME FEMALE CHARACTER.

Her identity must remain completely stable in EVERY FRAME.

This includes:

normal frames

fast-motion frames

motion blur

transition frames

pixel transformation frames

clothing-change frames

partially occluded frames

close-ups

profile angles

extreme angles

frames where another person crosses in front of her

Never restore the original woman's identity.

Never reveal the woman's face.

Never reveal the woman's body.

Never blend the reference video female character with the woman of reference pictures.

Never create a temporary female version.

Never morph from male to female and back.

Even if the transformation lasts only a few frames, those frames must still contain the SAME female character. @Image 2

==================================================

MOST IMPORTANT OUTFIT CHANGE RULE

THE CLOTHING TRANSFORMS.

THE HUMAN DOES NOT.

Every pixel transformation, digital breakup, clothing replacement, color change, texture change, or outfit transition applies ONLY to the clothes.

The female character underneath remains physically identical.

Correct visual logic:

SAME female character wearing Outfit A

→ pixel effect begins across clothing

→ clothing breaks into / passes through digital pixels

→ SAME female character @Image 1 remains underneath

→ Outfit B forms on the SAME body

→ SAME female character @Image 2 continues moving

Never:

female character

→ pixels

→ original woman

→ clothing changes

→ female character returns.

The pixel effect must never act as an identity-change transition.

==================================================

PIXEL EFFECT BEHAVIOR

Recreate the original reference video's pixel effect as closely as possible.

Match:

direction

density

speed

size

timing

breakup behavior

digital fragmentation

reassembly

interaction with the body

interaction with the other man

moment when clothing changes

The pixels should feel integrated into the physical scene.

When the other woman passes through the pixel field, preserve that original interaction exactly.

Her body should visibly move through / be partially occluded by the pixels in the same way as the reference video.

Do not replace this with:

smoke

magic

particles unrelated to the reference

light beams

portals

generic dissolves

morphing bodies

Keep the distinctive pixel clothing-change effect from the source.

==================================================

FACE LOCK DURING PIXEL TRANSITIONS

The female character's FACE @Image 2 is permanently locked.

If pixels move across the face or temporarily obscure part of it, the visible portions of the face must remain the same identity.

When the pixels pass away, the exact same face must be underneath.

Never reconstruct hidden facial areas using the original woman.

The face before, during, and after every transformation is the same person.

==================================================

BODY LOCK DURING OUTFIT CHANGES

Keep the female character's body proportions stable continuously.

The clothes adapt to her @Image 2 body.

Do not reshape her body to match the woman from the source video.

Maintain:

female shoulder width

torso proportions

waist proportions

arms

legs

height

posture

hand size

overall silhouette

The original woman's body is motion guidance only.

The female character performs her movement while remaining physically herself. @Image 1

==================================================

MOTION MATCHING

Follow the reference video's @Video 1 movement as closely as possible.

Do not reinterpret the choreography.

Match:

exact body direction

exact turns

exact head movement

exact timing

exact pauses

exact walking movement

exact hand placement

exact interaction

exact acceleration

exact deceleration

exact movement through the pixel effect

The character replacement should feel like the same footage was filmed with a different actor.

Do not invent new gestures.

Do not simplify the original performance.

Do not change the rhythm.

==================================================

FACIAL EXPRESSION

Copy the original woman's facial-expression TIMING and emotional beats while keeping the replacement female identity @Image 2

Match:

gaze direction

eyebrow movement

smiles

attitude

reactions

head tilts

timing of expression changes

The expression may follow the source, but the actual facial anatomy must always remain the female character.

==================================================

OTHER WOMAN IN THE VIDEO

Keep the other woman from the original reference video unchanged.

Do not replace her.

Do not modify her identity.

Do not alter her expression.

Do not change her clothing unless the original reference itself changes it.

Preserve her:

exact facial expression

body motion

timing

position

interaction

movement through the pixels

reaction to the transformation

she should behave exactly as she does in the reference footage.

Only the original woman is replaced.

==================================================

CAMERA

Follow the reference camera movement exactly.

Preserve:

camera direction

camera speed

framing

tracking

rotation

zoom

distance

perspective

movement timing

transition timing

Do not redesign the shot.

Do not make the camera more cinematic.

Do not add extra camera movement.

Do not remove original camera movement.

The goal is maximum motion fidelity to the reference video.

==================================================

TIMING

The reference video's timeline is absolute.

Every major action must happen at the same relative moment as the reference:

character movement

interaction

pixel activation

clothing transformation

other woman's movement through the pixels

expression changes

camera movement

transition completion

Do not stretch the sequence.

Do not shorten the sequence.

Do not insert extra pauses.

Do not move outfit changes earlier or later.

==================================================

TRANSFORMATION CONTINUITY

During every outfit transformation:

Frame before:

female character wearing current outfit.

Transformation frames:

SAME female character remains present while pixels modify ONLY the clothing.

Frame after:

SAME female character wearing the new outfit.

There must be continuous identity across all three stages.

Imagine tracking the character's skull, face, skin, hands, arms, body and silhouette through every transformation frame.

Those physical features must remain locked.

Only fabric, clothing color, clothing shape and clothing texture are allowed to change.

==================================================

OCCLUSION RULE

When pixels or another person cover part of the female character, do not use that occlusion as an opportunity to regenerate the original woman.

Hidden areas must remain mentally anchored to the same male body underneath.

When the occlusion clears:

reveal the SAME female character @Image 2.

Not the reference woman.

This applies especially when:

the other woman crosses in front

the pixel cloud becomes dense

clothing changes rapidly

motion blur increases

the camera changes angle

==================================================

REALISM

Keep everything photorealistic and physically believable.

Use:

realistic skin texture

natural hair

realistic clothing fabric

believable fabric folds

realistic shadows

proper contact between clothing and body

natural motion blur

correct lighting interaction

consistent depth

realistic pixel occlusion

Do not make the character look:

cartoon

anime

plastic

over-sharpened

game-rendered

waxy

==================================================

DO NOT CHANGE THE ORIGINAL EFFECT

Preserve the visual concept of the reference exactly.

Do not replace the pixel effect with a different transformation style.

The important visual moment where the other man passes through / interacts with the pixel clothing-change effect must remain.

Maintain the same sense that the digital pixels are occupying physical space and the other man is moving through them.

==================================================

FINAL PRIORITY ORDER

If Seedance encounters conflicting reference information, follow this order:

FEMALE @Image 2 CHARACTER IDENTITY

REFERENCE VIDEO MOTION + TIMING

REFERENCE VIDEO CAMERA MOVEMENT

REFERENCE VIDEO PIXEL / OUTFIT EFFECT

REFERENCE VIDEO EXPRESSIONS

OUTFIT DETAILS

Identity must NEVER be sacrificed to recreate the transformation.

If reproducing the source effect perfectly would cause the original woman to reappear, slightly simplify the effect instead.

NEVER sacrifice identity continuity.

==================================================

FINAL FAILURE PREVENTION

DO NOT bring the original woman back during outfit changes.

DO NOT use the woman as an intermediate transformation frame.

DO NOT switch gender during pixel effects.

DO NOT change face when clothes change.

DO NOT change body when clothes change.

DO NOT let pixels transform the person's identity.

DO NOT alter the other WOman's expression or movement.

DO NOT change the original camera choreography.

DO NOT change the timing.

DO NOT replace the source pixel effect with generic magic.

The FEmale character is permanently locked from the first frame to the final frame.

OUTFIT CHANGES.

PIXELS CHANGE.

CLOTHING CHANGES.

THE PERSON NEVER CHANGES.

[Maria Angelena @Angelena143us]

Matrix-level action meets cyber-pop style underground.🕶️⚡ Flawless multi-subject consistency, neon reflections, and gravity-defying motion. Generated with @ImagineArt_X @imagineart_creo #AIFashion #MariaAngelena #DigitalStyle

[Anissa @SimplyAnnisa]

A little glimpse of my India travel vlog

Made with seedance 2.5

Create a 30-second 16:9 photorealistic smartphone travel vlog featuring one Korean woman in her early 20s experiencing Agra, India for the first time. Make it feel like genuine handheld phone footage: spontaneous, immersive, slightly imperfect, natural lighting, realistic skin texture, autofocus, exposure changes, motion blur, walking jolts, authentic Indian ambience and real-world physics. No cinematic/commercial look, beauty filters, drone or gimbal shots.
Character: Same Korean woman throughout, petite oval face, brown eyes, natural coral lips, realistic fair skin, long black hair. No identity drift, face morphing or duplicates.
0:00–0:15 Outfit: Cream ribbed top, dark brown pleated mini skirt, tan belt, white knee-high socks, brown loafers, pearl earrings, loose half-up ponytail.
0:15–0:30 Outfit: Ivory embroidered sleeveless kurta, cream palazzo trousers, peach-orange chiffon dupatta with subtle gold embroidery, gold jewelry, nude sandals, thick loose braid over left shoulder. Hard cut exactly at 0:15.
STORY:
0:00–0:03: Taj Mahal first reveal. Selfie → turns camera toward monument → genuine awe. Korean dialogue: “와… 진짜 타지마할이다.”
0:03–0:06: Handheld Taj Mahal details, garden, reflecting pool and distant visitors.
0:06–0:09: Busy old Agra street with rickshaws, scooters, shops and pedestrians. She laughs and says: “여기 진짜 정신없다.”
0:09–0:12: Local bedai/kachori stall. Fresh food, steam, crispy texture. She tastes it and says: “어? 맛있는데?”
0:12–0:15: Walking toward Agra Fort, showing its massive red sandstone walls. She says: “와… 진짜 크다.”
0:15–0:18: Sadar Bazaar in new outfit, colourful shops and natural street activity.
0:18–0:21: Local petha shop. She tastes petha and says: “생각보다 엄청 달아.”
0:21–0:24: Roadside chai stall. She drinks hot chai with visible steam. Natural ambience, no dialogue.
0:24–0:27: Mehtab Bagh at golden hour. Taj Mahal realistically distant across the Yamuna. Gentle breeze moves her dupatta. Quiet genuine awe.
0:27–0:30: Agra street at night. Warm shop lights, traffic and pedestrians. She smiles and says: “아그라, 다음에 또 올게!” Then covers the lens with her hand.

[𝗦𝗮𝗻𝗶𝗮 @saniaspeaks_]

Made with seedance 2.5

Prompt
Extremely realistic smartphone travel vlog featuring one Korean woman in her early 20s visiting a modern zoo for the first time. The entire video must look like genuine footage casually recorded on a real smartphone by a friend and occasionally by herself, with imperfect handheld framing, natural phone-camera shake, realistic autofocus hunting, automatic exposure adjustments, slight motion blur, natural walking jolts, realistic skin pores, individual hair strands, authentic daylight, real shadows, natural reflections and physically accurate movement. No cinematic camera, no gimbal, no drone, no beauty filter, no artificial sharpness, no slow motion, no commercial look, no staged animal behavior.

Character: Same Korean woman throughout the entire video, petite oval face, brown eyes, natural coral lips, realistic fair skin, long black hair, natural makeup. She must remain perfectly consistent with no face morphing, identity drift, duplicate person or sudden changes.

0:00–0:04 — Zoo Arrival
She walks through the zoo entrance holding her smartphone in selfie mode. Real visitors naturally pass behind her. She smiles with genuine excitement and says in Korean: “안녕하세요! 오늘 동물원에 왔어요!” She briefly turns the phone toward the zoo pathway.

0:04–0:08 — Panda Encounter
She reaches the panda enclosure and quietly reacts when she sees a panda eating bamboo. She turns the phone toward the panda, then back toward herself. The panda behaves naturally and does not look directly at the camera. She softly says: “와… 너무 귀여워.”

0:08–0:12 — Giraffe Feeding
At a designated feeding area, she carefully holds a small piece of approved food. A giraffe naturally approaches and takes it from her hand. She laughs genuinely and instinctively pulls her hand back slightly. Keep the animal's movement realistic and anatomically accurate.

0:12–0:15 — Monkey Surprise
She walks past the monkey enclosure while filming. A monkey suddenly moves closer to the glass, surprising her. She takes a small step back, laughs naturally and says: “어? 너무 가까워!” Other visitors continue behaving naturally in the background.

0:15–0:18 — Outfit Change
Hard cut exactly at 0:15. Same woman, now wearing a light beige cardigan, white top, dark brown trousers, small crossbody bag and simple sneakers. She is walking through another zoo section. The outfit change must be clean and instantaneous while her face and identity remain identical.

0:18–0:21 — Elephant Encounter
She stops at the elephant enclosure. An elephant naturally moves its trunk while eating. She watches quietly with a genuine fascinated expression, then briefly looks toward the phone camera.

0:21–0:24 — Penguin Area
She crouches naturally beside the viewing glass while penguins swim underwater. One penguin swims quickly past the glass. She smiles and follows it with her eyes. Capture realistic reflections on the glass and natural indoor zoo lighting.

0:24–0:27 — Walking & Snack
She walks along a zoo pathway while holding a small ice cream. She takes one bite while continuing to walk, smiles naturally, and briefly shows the ice cream to the camera. Background visitors, trees, signs and distant animal sounds remain naturally active.

0:27–0:30 — Goodbye
She stops outside near the zoo exit in late-afternoon light. In selfie mode, she smiles naturally, slightly tired but happy after the day, waves toward the phone and says: “오늘 정말 재밌었어요! 다음에 또 올게요. 안녕!” She lowers the phone naturally, ending the recording.

[Maria Angelena @Angelena143us]

Pleated skirts, knee-high boots, and endless city views. 🏙️🤎 Flawless fabric textures, realistic wind motion, and stunning urban environments via @ImagineArt_X @imagineart_creo #AIFashion #MariaAngelena #DigitalStyle
What thing are you working on?

[Nailai7981 @VIBEQUIRKLABS]

Imagine 2.0 :
Photorealistic editorial portrait.

Knee-up cowboy shot, waist-level view, front view. Photographed directly from the front. Her chest and pelvis face the lens, with her torso, hips, and thighs forming a frontal knee-up silhouette.

A 20s seductive stunning Japanese or Korean woman with black frame, retro round-frame glasses, and small pearl stud earring detail, soft understated pearl accent, 94-58-92 body proportion anchor, full bust, narrow defined waist, rounded hips, dramatic but lean bust-waist-hip curve, young seductive alluring beauty face, Korean collarbone-length soft waves, defined layered shape, clean face-framing ends, lightly wind-swept movement, gentle directional flow, a few naturally loose strands, controlled overall silhouette, natural black hair, soft natural smile, relaxed brows, gently narrowed eyes, softly parted lips, lifted cheeks.

She wears fitted outerwear proportion, light blue denim jacket, washed denim texture, chest pockets, metal buttons, outerwear worn open at the front, front panels parted naturally, outerwear deliberately worn off both shoulders with both shoulder lines exposed and the upper back exposed in rear or three-quarter views, jacket draped around the upper arms with sleeves still on the arms, layered over, white strapless tube top, clean upper edge, smooth stretch fabric, compact bandeau or halter-bandeau structure, ultra-low-rise waistband sitting very low on the hips, standard lower-body proportion, off-white cargo shorts, compact utility cut, side cargo pockets, structured casual silhouette, black thigh-high sheer garter stockings, lace top band, visible garter straps, delicate nylon hosiery, decorative wide leather waist belt worn loosely around hips with a structured off-center buckle.

She has her head turned into a clean side profile with the face oriented away from the camera, and presents a casual, relaxed, and natural standing pose with one hip resting against an existing waist-height edge in the scene, clear hip-to-edge contact with partial body-weight support.

The scene is Dotonbori Shinsaibashi riverside edge in Osaka, iconic billboard signage, Shinsaibashi bridge railing in the foreground.

Deep azure summer sky, saturated clean blue atmosphere, hard direct sunlight on the subject, crisp shadow edges.

Rinko Kawauchi-inspired airy high-key image language.

Shot on 85mm short telephoto portrait lens, compressed facial perspective, flattering subject proportions, localized veiling flare, partial low-contrast haze around strong light entry, washed highlight patching.

Vintage sharp rendering, rich micro-contrast, subtle warm highlights, clean 35mm color density, precise edge definition.

[rui｜∫varts (シュバルツ) @varts_works]

MiniMax H3 — One character sheet, reshaping space.

キャラシート1枚から、空間そのものを操る15秒映像🏙️

指を動かすだけで、巨大なモニュメントが出現。
白・黒・��レーだけの世界で、空間全体が変形していく映像を作ってみました⚙️

制作: @TapNow_AI @TapNow_JP #TapNow #MiniMax

Prompt ↓

[Dua Fatima @DuaFatimaAi]

Gpt image 2 via chat gpt

Prompt

🎬 15-Second Dragon Fruit Juice Commercial Prompt

PROMPT:

Create a 15-second ultra-realistic cinematic food commercial in 9:16 vertical format, featuring a vibrant red dragon fruit transforming into a refreshing dragon fruit juice. Premium action-food advertising style, extreme macro photography, dramatic cinematic lighting, rich red and magenta colors, realistic fruit texture, dynamic camera movements, crisp sound-design moments, fast energetic cuts, and highly appetizing visuals.

Scene 1 — The Cut | 0–1.5 sec
Extreme macro close-up of a fresh red dragon fruit on a dark cinematic surface. A sharp stainless-steel knife rapidly slices through the fruit. Show the blade cutting cleanly through the vibrant pink flesh and green skin. Dynamic push-in camera movement, dramatic lighting, tiny juice droplets and realistic texture. Add the sound of a sharp knife slice and heartbeat-like impact.

Scene 2 — The Reveal | 1.5–3 sec
The dragon fruit separates into two halves, revealing its intensely vibrant magenta-red interior filled with tiny black seeds. Camera pushes smoothly between the two halves. Extreme macro detail, glossy moist texture, bright saturated colors, cinematic depth of field. Transition with a fast whoosh.

Scene 3 — The Extraction | 3–4.5 sec
Close-up of a spoon rapidly scooping the juicy dragon fruit flesh from the skin. Show thick, juicy pulp falling into a transparent blender container. Fast tracking movement, natural highlights, realistic seeds and pulp texture. Add subtle scraping and swooshing sounds.

Scene 4 — The Ingredient Attack | 4.5–6 sec
Rapid cinematic montage: dragon fruit pieces, ice cubes, milk and sugar drop dramatically into a clear blender jar. Ingredients fall from different directions in slow motion while the camera makes quick macro cuts. Bright clean lighting, crisp reflections, realistic splashes and ice movement.

Scene 5 — The Red Vortex | 6–8 sec
The blender switches on. Dragon fruit, ice and ingredients violently swirl into a spectacular deep-red vortex. Camera circles dynamically around the blender while the mixture spins rapidly. Thick creamy texture, realistic fluid simulation, glowing red highlights, energetic motion, powerful blender sound.

Scene 6 — The Red Transformation | 8–10 sec
The blender opens. A spoon lifts a thick, silky, intensely pink-red dragon fruit mixture. The smoothie slowly stretches and drips from the spoon in a beautiful glossy ribbon. Extreme macro shot, appetizing texture, shallow depth of field, slow cinematic push-in.

Scene 7 — The Red Pour | 10–12.5 sec
The vibrant dragon fruit juice is poured into a crystal-clear glass filled with ice cubes. Follow the stream with a smooth tracking camera. Show the thick red smoothie flowing around the ice, creating realistic bubbles, splashes and reflections. Premium beverage-commercial lighting.

Scene 8 — ONE CUT Hero Shot | 12.5–15 sec
Final hero shot of the finished dragon fruit juice in a tall transparent glass packed with ice. Place fresh dragon fruit halves around the glass. Camera performs a smooth 180-degree orbit with a subtle push-in. The drink looks extremely fresh, cold, glossy and irresistible. Finish with a cinematic fade-out.

Visual Style:
Ultra-realistic food photography, premium commercial cinematography, extreme macro lens, 85–100mm macro look, shallow depth of field, cinematic highlights, realistic liquid physics, high-speed photography, crisp textures, dramatic contrast, rich magenta-red palette, photorealistic ice, natural reflections, sophisticated advertising aesthetic.

Overall Direction:
Fast-paced, energetic, visually satisfying, curiosity → transformation → refreshing payoff. Every transition should feel connected and intentional, maintaining the same dragon fruit, glass, ingredients and visual identity throughout. No cartoon look, no CGI appearance, no distorted fruit, no unrealistic hands, no text or logos.

[TSUBAKI @AI__TSUBAKI]

Emma Asagiri 🇯🇵×🇬🇧×🐰

Made with Seedance 2.5 on @itsPolloAI
Generated in 1080p, then upscaled to 4K

Make sure to watch it in 4K~

[serein @you1873118]

提示词太长了

我做了一个SKILL

不想看那么多提示词

可以看下这个

https://t.co/OOEDEdjo5m

[Maria Angelena @Angelena143us]

Bringing futuristic streetwear straight to the dance floor. 🪩🛸 Flawless fast-motion tracking, seamless floorwork transitions, and realistic studio lighting. Created on @ImagineArt_X @imagineart_creo #AIFashion #MariaAngelena #DigitalStyle #imagineart

[アフガキくん @afugaki_sss]

【AI美女生成の救世主！？】

同じ人物を出力したい人は
細部までこだわった

「トリセツシート」

これを使ってみて。

詳細は以下にまとめた🔻

（保存推奨）

[leo @leomeethewoo]

AI will never be able to fully produce these videos end-to-end

Here's what it really takes:

1) Ideation: multiple people with years of editing and animation experience come up with ideas, pacing etc.

2) Design: each animation gets designed by multiple product designers

3) Production: professional motion designers work together on it & the art director makes sure everything hits the quality standard

Using AI as part of the process as we did in the example doesn't hurt

But that's 1% of the result, don't be fooled.

[Nawal @nawalsehar]

A simple ride, a quick grocery run, and the journey home.

Created with Seedance 2.5

Prompt :30-Second Cinematic

Ultra-photorealistic Korean summer lifestyle video following a young woman as she unlocks her city bicycle, places her crossbody bag in the front basket, and naturally rides through a peaceful residential neighborhood.

She slows down, turns realistically, parks beside the grocery store, locks her bicycle, and enters the supermarket. She casually selects a few groceries—apples, milk, and a snack—then pays at the checkout.

After placing the groceries securely in the bicycle basket, she unlocks the bike, mounts naturally, and cycles home along the same street.

Focus on realistic human movement, bicycle balance, pedal motion, braking, tire contact, momentum, wind, clothing and hair movement, grocery-bag weight, and authentic everyday sounds. Smooth cinematic camera movement, natural lighting, realistic depth of field, and continuous cause-and-effect action.

No teleportation, instant acceleration, floating objects, distorted anatomy, changing clothes or bicycle, exaggerated wind, CGI appearance, or action-movie movement.

[𝐒𝐊_𝐀𝐈 @Its_Saad_kkk]

Made with Gemini AI

Created a fresh visual campaign for Hey, Sugar! A sweet take on body care, blending a delicious peach-inspired aesthetic with a luxurious self-care experience.

Video Prompt 👇👇

Shot 1 Peach Product Hero
Show the Hey, Sugar! scrub jar beside fresh peach slices on a glossy peach-colored surface. Warm sunlight, soft shadows, and a slow cinematic push-in. Premium beauty campaign aesthetic.

Shot 2 Lifestyle Product Hold
A model holds the scrub jar close to her shoulder and chest. Natural glowing skin, soft warm sunlight, elegant minimal styling. Camera slowly moves from her face toward the product.

Shot 3 Texture Close-Up
Extreme macro shot of the open scrub jar, highlighting the creamy peach-colored sugar texture. Small rose petals and subtle grains surround the product. Slow focus pull across the texture.

Shot 4 Beauty Portrait
Close-up beauty shot of the model holding the scrub beside her face. Warm peach background, glossy skin highlights, confident natural expression, soft editorial lighting.

Shot 5 Application Demo
Close-up of the scrub being gently applied to the skin on the arm/shoulder. Capture the grainy texture and smooth application in realistic detail. Slow-motion movement with warm sunlight.

Shot 6 Opening the Product
Hands slowly twist open the Hey, Sugar! jar, revealing the fresh scrub inside. Clean peach backdrop, detailed packaging, soft shadows, premium product-commercial lighting.

Shot 7 Surreal Product Shot
The scrub jar balanced on top of a fresh peach against a warm peach background. Minimal composition, dramatic sunlight and shadow, slow camera orbit.

Shot 8 Final Beauty Hero
Model gently applies the scrub to her shoulder while the product remains visible in the foreground. Warm peach tones, glowing skin, soft cinematic lighting, luxurious self-care atmosphere.

Overall style: ultra-realistic beauty commercial, warm peach palette, soft golden sunlight, premium skincare photography, realistic skin texture, elegant compositions, smooth cinematic camera movements, consistent product packaging and branding.

[MrDejie @mrdejie]

20 shots in 30seconds.
Yes! you can do that much with Seedance 2.5!

Made on @lovart_ai

Prompt 👇
Create a 30-second cinematic continuous one-shot where both the world and the character are constantly moving. The transformation must never feel like a girl simply walking while backgrounds change. Her body, momentum, balance, position, and reactions should actively cause or motivate each transition. No cuts, no fades, no dissolves. The camera may whip, orbit, dive, roll, pass behind objects, move overhead, go low to the ground, rush toward her, or briefly lose sight of her behind foreground elements, but it must remain one physically continuous shot.

Character: A realistic Japanese girl in her early 20s. Natural features, long dark hair, athletic and expressive. Casual modern clothes that stay readable as they get wet, dusty, torn, and transformed by each world. Photoreal, not anime.

Start in a bright Candy World. Giant lollipops, frosting hills, candy-cane trees, and sugar dust in the air. The girl hears a rumble and spins around. The camera rushes backward in front of her as she runs toward a gigantic painting of a lava cake. She skids to a stop. Molten chocolate lava suddenly explodes out of the canvas. She dives sideways as real lava-cake flood crashes across the candy floor. The camera drops almost to floor level with the rushing chocolate. She grabs a giant candy bench to stop herself. The bench stretches beneath her hands and becomes a fallen log. The camera swings around her. She is suddenly in the Year of the Dinosaurs, hanging from a tree as a massive dinosaur charges past. She pulls herself up, runs, jumps over a sliding boulder, ducks beneath a swinging vine, then grabs a giant leaf. The camera rises with the leaf. The leaf wraps around her. She tears through it and tumbles onto a paper world. The same leaf is now a huge sheet of folded paper. She rolls across paper mountains, pushes herself up, and immediately begins running as paper buildings unfold around her. The camera races sideways with her. Paper edges whip past the lens and transform into glass. The paper city becomes a modern skyscraper corridor. She runs across glossy floors as windows race past. She loses her footing near an open shaft and falls. The camera dives with her. The glass walls become rough stone. She lands hard inside a prehistoric BC cave. Torches flicker. She crawls forward, then sprints as the cave opens onto an ancient battlefield. Spears and banners rush past. She drops flat as riders leap over her. From ground level a giant war banner falls toward the lens. She grabs its edge. The wind yanks her upward. For a moment she is lifted completely off the ground. As the camera rotates with her, the banner becomes the feathers of a gigantic bird. She is suddenly riding on the big bird’s back high in the sky. She holds on as it banks through clouds. The bird dives. She slips, grabs a talon, then drops. The camera falls with her. She plunges into an underground river. Now underwater, her momentum continues naturally. She swims with the current through a dark cavern, then breaks the surface onto an abandoned island. Skeletons lie scattered across the sand and rocks. She stands, panting, then steps on a glass shard. The island shrinks around her. The camera pulls back as the whole island is revealed inside a giant glass bottle. She is trapped inside the bottle. The bottle tips. She slides, slams against the glass, and the glass becomes a dusty window. She crashes through it into a creepy house. Creaking floors, peeling wallpaper, long hallways. She runs. A door slams open. The house corridors fill with a zombie world. She dodges reaching hands, leaps over fallen furniture, and bursts through a classroom door. She is suddenly in a school classroom. Desks scatter as she sprints across them and jumps through an open window. The camera follows her through the window. She does not land in a yard. She lands on the top of a building. Wind slams her. She runs across the rooftop, then slips off the edge. The camera dives beside her. She hits water. Now underwater again. She swims upward toward a bright shape. When she breaks the surface, she is pulling herself into a pilot’s cockpit of a flying plane. Instruments shake. She grabs the controls, then the plane punches through clouds into a dense forest. The cockpit becomes tree branches. She tumbles through leaves, rolls across moss, and slams onto a mattress.

Instantly she is sitting upright in bed. Morning. Silence. The camera is now extremely close to her face. She gasps awake. For the first time in the entire film, everything becomes completely still. Her bedroom is ordinary. She looks around. She touches the mattress. Checks her hands. Relieved, she lies back down. Then something drips onto her forehead. She looks up. The ceiling above her bed is gently rippling like candy glaze mixed with water. A large shadow slowly moves across it. Her expression changes from relief to disbelief. Then the entire bed suddenly tilts. She grabs the mattress as her alarm clock, pillow, lamp, and furniture begin sliding toward one side of the room. End exactly as she begins sliding out of bed, suggesting that waking up was only another stage of the transformation.

The action should feel choreographed like a physical adventure. The girl must run, sprint, skid, dive, fall, climb, hang, swing, crawl, jump, roll, swim, land, sit, get thrown through spaces, and briefly become weightless. Her position within the frame should constantly change: close-up, full-body, profile, silhouette, overhead, low-angle, behind her, directly in front of her, underwater, upside down, falling beside her, and orbiting around her.

Transitions should be triggered by her actions. When she grabs something, it becomes something else. When she falls, the next world catches her. When she jumps through an opening, that opening becomes another location. When an object crosses the camera, it transforms while hiding part of the environment.

Use physical match transitions such as:
candy bench → dinosaur-era log
giant leaf → paper sheet
paper edges → skyscraper glass
falling shaft → BC cave stone
war banner → giant bird feathers
bird dive → underground river
island sand → bottle glass
bottle glass → creepy house window
house door → zombie world
zombie corridor → classroom
classroom window → rooftop
rooftop fall → underwater
water surface → pilot cockpit
cockpit / clouds → forest branches
forest ground → bedroom mattress

The camera movement itself should help generate transitions. A whip-pan can turn one moving object into another. A 360-degree orbit can reveal a completely transformed environment by the time it completes. Passing behind fabric, pillars, people, waves, vehicles, doors, or the girl’s body can briefly obscure parts of the frame while the transformation happens continuously.

Despite the speed, every transformation must remain visually understandable. The result should feel less like a transformation montage and more like one impossible 30-second physical journey in which the girl is being chased, thrown, carried, dropped, and pulled through realities that continuously grow out of one another.

Tone: exhilarating, mysterious, playful, cinematic, photoreal, large-scale, dreamlike, and full of wonder.